Inferring causal relationships among growth curve traits of Lori-Bakhtiari sheep using structural equation models
Introduction
Lori-Bakhtiari sheep is one of the most important Iranian native sheep breeds in southwestern of Iran, with a population of more than 1.7 million heads (Vatankhah et al., 2019). This breed is more pronounced for meat production and lambs with fast growth especially from birth to six-month of age.
Growth is one of the most important characteristics of livestock animals (Bathaei and Leroy, 1998; Keshkin et al., 2010; Ghavi Hossein-Zadeh, 2015c, 2017) and can be described as an increase in live bodyweight over time. Moreover, changes in live bodyweight across time are explained by the growth curves (Keshkin et al., 2010). Analysis of the growth performance of animals throughout their lifetime is a valuable tool for administration of the appropriate feeding method and selection for attaining the best slaughter age (Ghavi Hossein-Zadeh, 2015c). Early body weights especially birth weight had a significant effect on latter body weights. Growth curve parameters may be used as criteria for changing the association between live body weight and age via selection and breeding (Kachman and Gianola, 1984). Furthermore, the growth curve parameters are heritable in different species and it may be possible to change the shape of growth curve through selection of animals for those parameters (Kachman et al., 1988). Slow growth rate resulting in low market weight has been identified to be one of the factors limiting profitability in any production system of animals (Noor et al., 2001; Abegaz et al., 2010).
The application of mathematical growth models provides an efficient method for summarizing the information included in a dataset into a few parameters to facilitate both the understanding of the growth trajectory and its interpretation (Malhado et al., 2009). In biological systems, phenotypic traits may have causal relationships but standard multivariate models (SMMs), which are used commonly for genetic evaluation purposes, cannot handle such causal relationships and only estimate association among the traits, while association cannot show causal relationships among the traits. Structural equation models (SEMs) consider simultaneous and/or recursive causal relationships among the traits (Gianola and Sorensen, 2004).
Genetic parameters of growth curve traits in several sheep breeds have been reported using SMMs (Ghavi Hossein-Zadeh, 2015b, 2017; Nimase et al., 2017). These studies did not take the potential causal relationships among the growth curve traits into account. A few numbers of studies have investigated the causal relationships between growth traits in small ruminants (Mokhtari et al., 2018; AmouPosht-e Masari et al., 2019), but these studies did not take into account the growth curve parameters. Growth curve parameters are related to rate of maturing and mature weight; furthermore, these traits have been suggested to have an association with lifetime productivity parameters (Pala et al., 2005). Finding the causal relationships between birth weight and growth curve parameters provides a sound basis for developing a breeding strategy to modify the trajectory of growth (Abegaz et al., 2010). Therefore, the present study aimed to infer the potential causal effects of birth weight on growth curve traits in Lori-Bakhtiari sheep using SEM, as well as to assess the potential causal relationships among the growth curve traits. Besides, estimates of genetic parameters, goodness of fit and predictive ability of SMM and SEM were compared.
Section snippets
Flock management
The flock was managed under a semi-migratory or village system. From December to May the flock was usually kept at the breeding station and was fed lucerne, barley, and wheat stubble indoors. The rest of the year the flock grazed on the range and cereal pasture. The mating period extended from late August to late October (20–25 ewes were assigned randomly to one fertile ram) and consequently, lambing started in late January and lasted to late March. Lambs were suckling from their dams until 15
Results and discussion
The goodness of fit statistics including R2adj, AIC, RMSE and DW for comparing the non-linear models describing the growth trajectory of Lori-Bakhtiari lambs are presented in Table 3.The Brody model supplied the highest values of R2adj (0.89) and DW (1.62) and the lowest values of AIC (149895.68). The estimates of growth curve parameters using the non-linear models are shown in Table 4.
The BIC values across the traits under six different animal models are shown in Table 5. The best animal model
Posterior means of direct heritability estimates
Posterior means of direct heritability of BW under SMM (0.32) and FRM (0.31) were in line with those reported by others (Safari et al., 2005; Tariq et al., 2010). Under both SMM and FRM, the estimates of direct heritability for b, k and A were 0.14, 0.16 and 0.10, respectively. Hence, the existence of potential causal relationships among the studied traits did not have significant effect on the direct heritability estimates of studied traits. Ghavi Hossein-Zadeh (2015a) reported values of 0.15,
Conclusions
The obtained results showed that the differences between direct heritability estimates from SMM and FRM were negligible, while correlation estimates between the studied traits using SMM and FRM were statistically significant. The causal effect of BW on k and A, showed that any improvement in the initial weight of Lori-Bakhtiari lambs has a positive effect on their latter weights. Therefore, BW is an important trait for improving the growth performance of Lori-Bakhtiari breed. However,
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of Competing Interest
The authors report no declarations of interest.
Acknowledgment
The authors would like to thank all the staff of Lori-Bakhtiari Shooli Station for data collection and the National Animal Breeding Center (ABC) of Iran for providing the data.
References (37)
- et al.
Genetic and phenotypic aspects of the growth curve characteristics in Mehraban Iranian fat-tailed sheep
Small Rumin. Res.
(1998) - et al.
Growth curve in Santa Ines sheep
Small Rumin. Res.
(2012) - et al.
Exploration of relationships between claw disorders and milk yield in Holstein cows via recursive linear and threshold models
J. Dairy Sci.
(2008) - et al.
Analysis of fertility and Dystocia in Holsteins using recursive models to handle censored and categorical data
J. Dairy Sci.
(2007) - et al.
Growth curve in Dorper sheep crossed with the local Brazilian breeds
Morada Nova. Rabo Largo. and Santa.SmallRumin. Res.
(2009) - et al.
The application of recursive multivariate model for genetic evaluation of early growth traits in RaeiniChasmere goat: a comparison with standard multivariate model
Small Rumin. Res.
(2018) - et al.
A review of genetic parameters estimates for wool, growth, meat and reproduction traits in sheep
Livest. Prod. Sci.
(2005) - et al.
Genetic parameters of body weight and fat-tail measurements in lambs
Small Rumin. Res.
(2008) - et al.
Population structure of Lori-Bakhtiari sheep in Iran by pedigree analysis
Small Rumin. Res.
(2019) - et al.
Estimation of genetic and phenotypic parameters of growth curve and their relationship with early growth and productivity in Horro sheep
Arch. Tierz.
(2010)